Date

Attendees

Goals

  • PGx team needs a Result Type variable and the Lab team needs a Result Qualifier variable.  Are they the same?  Then both teams can join together to request a new variable.

Discussion items

TimeItemWhoNotes
   

PGx team:  draft proposal for a "result type" (result format) variable to meet the needs of the Lab terminology and PGx teams.

     - SDTM model (--RESFMT):  This variable is used to specify the format of the result; i.e., what the result is reported as.

     - LB domain (LBRESFMT):  This variable is used to specify the format of the result; i.e., what the result is reported as. Examples: RATIO, INTEGER.

     - PF domain  (PFRESFMT):  This variable is used to specify the format of the result; i.e., what the result is reported as. For genetics/genomics, this variable indicates the molecular characterics of the result. Examples: NUCLEOTIDE, AMINO ACID, HAPLOTYPE, * ALLELE, HUMAN LEUKOCYTE ANTIGEN

   

Lab team:  draft proposal for a "result qualifier" variable

     - Existing lab test "Glucose-6-Phosphate Dehydrogenase"  (C80184)         Definition:  A measurement of the glucose-6-phosphate dehydrogenase in a biological specimen.

     - New request to add "Glucose-6-Phosphate Dehydrogenase Activity".       Definition: A measurement of the glucose-6-phosphate dehydrogenase activity in a biological specimen.

     - The lab test codelist contains the analyte.  The specimen, method, and unit variables give further definition to the analyte.  Concentration and Activity will have different methods and units but have the same analyte.

     - One lab test for Glucose-6-Phosphate Dehydrogenase and the new Result Qualifier Variable could contain ACTIVITY. 

   

Test Qualifier or Result Qualifier:

PGx examples are result qualifiers/result types.

Lab examples:

Fred - this is a test qualifier.

Lab team - Concentration that is mass, a specific activity, or plain concentration.  These are all result types.  They are telling you what the thing is that is being recorded.  It is a test qualifier that explains the result.

     - it is a test qualifier that tells you what the result is.  So it should be stored in the Result Qualifier variable instead of the Test Details.

     - Example for test details:  Genetox - this was run under conditions where it had no metabolic activity added to it or it did have metabolic activity added.  This would belong in test details/test qualifier.

   

Diane had been at a previous meeting with the lab terminology team on 03Aug2016 to discuss the use of the Test Details variable.  At that meeting, the team also discussed the terms from the KT TAUG  http://wiki.cdisc.org/x/sR-UAQ  and had decided that these were not test details but result qualifiers. 

Example 6,7,8 "from the Anti-HLA Antibody Screening section in this TAUG demonstrates a situation wherein the test will yield 3 results referencing the same antibody: the Presence/Absence, the Mean Florence Intensity (MFI), and the Titer of the antibody. Following CDISC’s LB domain rule, the LBTEST = XXX Antibody; this LBTEST appears in 3 separate rows with 3 separate results: Presence/Absence of the antibody, MFI value, and Titer value. The CDISC LAB controlled terminology team suggested that both MFI and Titer are different types of results. Specifically, they are result qualifiers and NOT TEST qualifiers, and as such they should not be represented by the LBTSTDTL variable. A new variable “Result Type” has been proposed, and subsequent to its approval the following values are candidates for this variable: MFI, Titer, Ratio, Mean Channel Shift, Virtual Crossmatch, Actual Crossmatch, etc." 

   

Example from Jordan KT TAUG:  a single lab test generates 3 results in the forms of presence/absence of an analyte, the MFI, and Titer of the said analyte  http://wiki.cdisc.org/x/sR-UAQ

Jordan comment: "The example in the KT-TAUG is the following: a single lab test generates 3 results in the forms of presence/absence of an analyte, the MFI and Titer of the said analyte. Based on the LB domain rules, the analyte itself is always the LBTEST, hence we have 3 separate records of the same LBTEST = XX Analyte and 3 results where LBORRES = Presence/Absence; LBORRES = xx numeric value (in MFI); LBORRES = xx numeric value (in Titer). The problem here is how to sufficiently differentiate the 3 records given that the LBTEST is the same in 3 rows.  The values in question “Detection, Titer, Mean Fluorescence Intensity (MFI), Mean Channel Shift, Virtual Cross Match and Actual Crossmatch” were first submitted as candidate values for LBTSTDTL, but the lab team objected based on the reason that these values are different ‘result types’ hence should be represented by a result qualifier variable."

   

Test Qualifier or Result Qualifier:

Fred - they are actually additional qualifiers of the test (PerformedObservation), and not part of the result. One knows before one runs each of the measurements, which result they will get before the test is run. These are not qualifiers of the result; in other words, one does not get a result, and then figure out what kind of test was run.

Lab team - they are result qualifiers.

   

No resolution during this meeting. 

Another meeting is needed.

Everyone is to review the BRIDG model before the next meeting.

Action items

  •